Freelance AI Data Contributor & Evaluator (Remote)
Evaluated AI prompts and rated model outputs for accuracy, helpfulness, tone, and safety according to provided guidelines. Contributed structured written feedback intended to support RLHF (Reinforcement Learning from Human Feedback) pipelines and improve response quality. Maintained consistently high quality scores while following multi-step annotation instructions across projects. • Rated AI responses on defined quality criteria (accuracy, helpfulness, tone, safety). • Provided RLHF-style written feedback to guide improvements in model behavior. • Followed detailed, project-specific annotation guidelines and quality scoring rubric. • Completed high-volume evaluation tasks within tight turnaround windows while maintaining compliance.